Lurking in the Sewer

Managing a collection system is a difficult task because everything about a collection system, except failure, is invisible. The challenge for the industry is to bring visibility to the events that reduce capacity and increase flow before one or both can result in outright failure.

A sewer is considered to have failed when certain events become visible, such as basement flooding, roadway collapses and, sanitary sewer overflows (SSOs). Absent these failures, many managers will assume that everything seems OK. Hydraulic models are popular because they remove much of the mystery by simulating and displaying what is likely to be happening in the sewer.

In reality there is a long list of issues that are invisibly changing the performance of a collections system. It is safe to say that no natural change in a sewer ever results in an improvement of performance. Every change, except human improvements, gradually causes deterioration in performance until visible signs of failure appear.

Figure 1 depicts sewer conditions and separates them into visible and invisible categories. The conditions lurking below the line of visibility are constantly working to deteriorate the system. These conditions will become visible when they combine to cause a failure, e.g., a worsening blockage combined with a rainfall of just-critical intensity. What is needed for sewer systems is an "indicator of change" that can bring attention to the problems lurking in all collection systems well before they make themselves visible.

A common source of frustration for anyone using flow data has been the delay between the moment an actionable event occurs in a sewer and the time a manager becomes aware of it by looking at flow data. Flow data could alert the manager to a lurking problem, but the data is generally delivered in weekly or monthly reports. In the past, flow data was collected on a laptop or desktop and data was available to only the single user. Data was made available by hardcopy reports. Staff analysts and consultants are used to having to sort through the data to find signs of overflow, surcharging blockages, or other actionable events. These analyses could allow the manager to direct day-to-day or near-term activities, yet the analyses may not reach the manager's desk for weeks or months.

A common example of a performance deficiency, due to lurking problems, that contributes to sewer failure is "parasitic hydraulic loss." Parasitic hydraulic losses are apparent at higher flow rates and are the result of construction deficiencies like poorly constructed manholes and turning structures. Something as simple as low manhole bench walls (half-pipe or less) can easily reduce a pipe's capacity by 30 percent. Parasitic hydraulic losses are generally not visible unless flow monitor data are examined in a scattergraph form as discussed below. Reverse grade sewers, even for just a few pipe segments, can rob a pipe of 20 percent or more of its peak capacity. Cinder blocks and other debris thrown in sewers by vandals may also reduce capacity such that a 2-year rain now causes surcharging and basement flooding when the same sewer previously could handle a 10-year storm with ease. Data from a nearby flow monitor could have shown the change in flow pattern almost immediately.

Today, early invisible changes, like these "lurking problems" can be detected with accurate flow-monitoring data combined with timely and appropriate analysis and delivered in real-time to municipal and collection system managers. In determining what type of flow monitoring your system requires, it is important to note that flow monitors deliver two distinct types of information: 1) time series data or hydrographs of calculated flow, which reveal information about the quantity of flow generated by the system upstream of the monitor; and 2) scattergraphs of depth and velocity data, which reveal both the hydraulic conditions in the pipe and how well the monitor itself is working. An analysis of time-series data for flow and rain will determine how much wastewater and RDII (rainfall dependent infiltration inflow) is being generated in each sewer shed, and the scattergraph analysis will determine the operational capacity of the sewer at each sewer shed. Together, these analyses determine how a system performs as envisioned by the pending CMOM program. However, the U.S. Environmental Protection Agency (EPA) often discusses the concept of using Performance Indicators and suggests that events, such as complaints, basement flooding, and overflows, be considered as measurable performance indicators. Many cities use this approach to determine how well their programs are working, but these measures are actually frequency-of-failure indicators, not true performance indicators. From an engineering perspective, the combination of an RDII analysis and a scattergraph analysis can reveal many of the lurking problems in a collection system. If performed regularly, these two analyses can identify "indicators of change" in sewers.

One obstacle to overcome for timely analysis is to collect near real-time data and make it available on a common platform to multiple users. The combination of wireless communication and the Internet make this much easier to accomplish, and the major flow monitor manufacturers are incorporating these technologies into their products. Telog Instruments has developed Telog Enterprise™ software, which can collect data from several types of flow meters and make the data accessible on the Internet. ADS Environmental Services has developed Web-based IntelliServe™ for automatic data collection and real-time alarming, as well as Sliicer.com™, for performing sophisticated engineering analysis. Data from any flowmeter is accessible over the Internet through these software products.

The market has seen a maturing of the Internet-accessible offerings. The earliest entry in the market combined the appeal of Internet data access with ineffective metering technology. The market soon realized that acquiring poor flow data over the Internet did not increase its value. Other manufacturers are in the process of developing other methods for users to acquire data over the Internet. Some methods simply allow a user to control desk-top software remotely over the Internet. However from the user's perspective receiving data more quickly and more conveniently does not by itself make it easier to spot the lurking problems in sewers. Still an additional level of software analysis is needed.

Engineering Analysis of Flow Quantities and Hydraulic Conditions

One of the most frequent analyses performed is the dry and wet weather analysis of flow -- often to determine levels of RDII in a sewer system. One of the goals of such work is to correlate rainfall with RDII in such a way that this rainfall-to-flow relationship is used to characterize the performance of a sewer basin. Capitol planning and sewer rehabilitation is guided by the severity of the rainfall-to-flow relationship of each basin within a collection system. Being able to monitor the rainfall-to-flow relationship for each basin is key to getting an early warning that an invisible problem is lurking.

Rainfall Analysis

Every collection system manager has a good idea of the rainfall needed to cause problems in their system, but some rains just don't follow the rules. "That 1.7-inch rain we had yesterday sure caused unexpected problems". This ADF display (Accumulation Duration Frequency) allows the manager to immediately "see" how the storm behaved. In this case the graph reveals that what appeared to be a small 2-year storm when looking at the storm total was actually a 30-year, 2-hour storm. Knowing this information can help explain why the sewer responded poorly, and it can even become a defense in an enforcement action. With the help of software, this analysis can be performed in seconds for each gauge and for each storm.

Scattergraph Analysis

The scattergraph is a great "human viewing speed" graphic that can reveal the hydraulic conditions in a sewer. Every SSO must be accompanied by a downstream restriction, whether it is a pump station, a treatment plant, or a problem with the pipe itself. A scattergraph can quantify the restriction and often can determine the type of restriction. This scattergraph is equipped with iso-Q lines (lines of constant flow rate) and shows that this pipe actually carries 60 percent of its design capacity and is surcharged to a depth of 70 inches. This operational capacity of a pipe can change over time and being able to quantify operational capacity is key to spotting "lurking problems." This problem could be the result of tree roots or a pending sewer failure. Note: For more information on Scattergraphs and how they can help us identify 'lurking problems' check out the information at www.adsenv.com/scattergraphs.

Q-to i Relationships

The rainfall-to-flow relationship is the key performance indicator of the wastewater production side of the collection system. It measures the "yield" or the amount of I/I generated in a basin. Tracking this yield can reveal one of the first "indicators of change" that a manager may see in the system. The manager of the Longmont Creek system was able to spot a serious problem in the MS04 basin by analyzing the Q-to-i relationship for just the first four storms in 2004 as shown in Figure 2. The dramatic shift indicates that new I/I was showing up in this basin. Investigation quickly discovered that a road construction contractor had incorrectly tied a storm sewer into the sanitary. Many collection system managers may sheepishly admit that such an occurrence could continue for years without being discovered.

Managers also rely on this rainfall-to-flow relationship to quantify the effect of rehabilitation. All too often, managers try to compare I/I volumes before and after rehabilitation by selecting one or two storms that are similar in magnitude. Inevitably there are other variables, such as antecedent rain or a different season of the year that interfere with the comparison. These Q-to-i relationships for several storms before and after make this evaluation much more effective.

This article originally appeared in the September/October 2005 issue of Water and Wastewater Products, Vol. 5, No. 5.

This article originally appeared in the October 2007 issue of Occupational Health & Safety.

Featured

Artificial Intelligence